Jiarui Fang bef05489b6 [zero] bucketized tensor cpu gpu copy (#368)

2022-03-10 14:41:08 +08:00

..

init_ctx

[zero] able to place params on cpu after zero init context (#365)

2022-03-10 14:08:58 +08:00

shard_utils

[zero] able to place params on cpu after zero init context (#365)

2022-03-10 14:08:58 +08:00

sharded_model

fix grad shape

2022-03-10 12:06:42 +08:00

sharded_optim

[zero] bucketized tensor cpu gpu copy (#368)

2022-03-10 14:41:08 +08:00

sharded_param

[zero] bucketized tensor cpu gpu copy (#368)

2022-03-10 14:41:08 +08:00

__init__.py

added buffer sync to naive amp model wrapper (#291)

2022-03-02 16:47:17 +08:00